Totalt antall sidevisninger

fredag 10. januar 2014

Euroasian variation gradiation maps

This is a graphic presentation of the 7 Euroasian haplotype based (chunkcount) PCA plots from the latest project run using 289k linked SNP's. The average number of SNP in each segment chunk in the world panel is 13 (with a heavy overweight of Europeans and Northern Europeans in particular). This analysis is at this stage experimental. Please ignore coloring in Africa, Australia and Greenland as no populations are included from these continents.

The first dimension peaks on one side among Finns and Saamis (brown), and on the other side among Panya, Chukchi and Cambodians (light blue or blue). It appears all Fennoscandians in general belongs to this brown component.. This component appear identical to dimension 1 in the previous analysis. As this is the first dimension it also explain the largest variation in haplotypes between the populations. As the "blue side" here appear mostly at the coast peaking among populations that show affiliation to Papuans and Melanasians I suspect it seperate this ancient population from old European hunter gatherers.

Euroasian dimension 1

The second dimension peaks on one side among Lithuanians and Scandinavians and on the other side among the "The Others" containing the remaing individuals, but in the remaining panel it is peaks among Miao and some other East-Asian population. It seem to show a clear division of West and East Euroasian populations. All Fennoscandians belong in general to this western cluster.

Euroasian dimension 2

The third dimension peaks on one side among the "Others" and secondary at Bedouins and other Middle East populations and on the other side among North Siberian populations. It seem to represent influence from the African continent. This influence have reached as far north as South-Scandinavia but to less extent among Saamis and Finns. This dimension may be related to dimension 3 in the previous analysis.

 Euroasian dimension 3 

The fourth dimension peaks among Dai, Cambodian and Han in South-East Asian on one side and among Koryak, Yugagir and Nganassans in North-Siberia. Saamis, Finns and to a degree Scandinavians seems most similar in variance to the North-Siberian group while Continental-Europeans appears more similar to the South-East Asian group.

 Euroasian dimension 4

The fifth component peaks among Saamis, Finns and some South-East Asian populations on one side and among Northern Siberians on the other. Scandinavians appear less related to this dimension. This dimension may be related to dimension 3 in the previous analysis.

This component appears more difficult to explain as other analysis have shown no connection between Saamis, Finns and South-East Asians but to North-Siberians as in dimension 4. It may be an effect of having large sample of Europeans and small sample of other populations however its striking that the clustering appears consistent among the various non-European populations and not spread out randomly as if there was no structure. As far as I know and can remember there has not been done such wide scale analysis before using linked haplotypes so it may be something not seen before.

  Euroasian dimension 5 

The sixth dimension peaks among East-Asians on one side and the Indian subcontinent on the other. Saamis  Finns appear a little closer to the Indian subcontinent than Scandinavians.

 Euroasian dimension 6  

The seventh dimension peaks among the Lithuanians and Koryak on one side and among the Chukchi, South-Indians and among some North-Siberians. In more general terms as the heatmap shows similarity between Eastern Europe and South-East Asians. This dimension as dimension 5 appears difficult to explain and the same stated about this there apply here as well.

 Euroasian dimension 7

The remaining higher dimensions appear to show local variation between Siberian groups. 

1 kommentar:

  1. The possibility of some kind of old East Eurasian-like component extending some way south and west of Fennoscandia seems to be there.

    It's really hard to explain it with only a component like Mal'ta ANE from Lazaridis et al, because in a test from that study all north europeans show not only a shift towards Mal'ta (Y-axis) but towards Han Chinese, who have shown no ANE ancestry, (X-axis) compared to the early european Stuttgart farmer. For example Lithuanians, Belorussians and Scots are closer to Mordovians (who are known to have old Siberian admixture) than to Stuttgart in their "Han-shift".